problem running 64-bit Linux version

[ssybesma]

Right now I am having a problem starting over.

I cleared the two jobs out of my queue (unreserved them) for this CPU (STCK1A8LFC-A079AF). I'm having trouble getting a new LL assignment.

I just want to do first checks on ONE new potential prime on this thing. The command line menu is confusing sometimes.

==========

UPDATE:

This thing is so unruly...I want to select an exponent (332375291) and it keeps sticking it under manual testing when I need it to show up under the machine that's doing the testing.

I have to have control of this and I don't yet.

[ssybesma]

How do I get rid of the manual testing and have that go under my Intel Compute Stick?

ssybesma has 4 registered computers

STCK1A8LFC-A079AF Intel Atom Z3735F @ 1.33GHz Linux64,v29.4,build 8 1567 2019-03-22 08:59 recently seen, trusted software T 0.0 LL-100M

Manual testing 2019-03-22 08:49 recently seen, untrusted software U 0.0 332,375,291 LL 0.0% 2019-03-22 08:57

NUC Intel Core i7-5557U @ 3.10GHz Windows64,v29.4,build 8 3382 2019-03-22 06:37 recently seen, trusted software T 245.7 LL-100M 332,819,869 LL LL, 17.9% 2019-06-16 06:02 2019-03-22 06:37 77 86

MAC Intel Core i5-4278U @ 2.60GHz Mac OS X 64-bit,v29.4,build 7 2883 2019-03-21 22:39 recently seen, trusted software T 1,697.1 LL-100M 332,229,923 LL LL, 10.6% 2019-06-20 20:44 2019-03-21 22:39 91 90

[ATH]

In local.txt check for the lines WorkerThreads and CoresPerTest and change them to:
WorkerThreads=1
CoresPerTest=4

In worktodo.txt change to only 2 lines:

[Worker #1]

and then the worker line for the test you want to keep.



To unreserve the exponent on the server go to:
https://www.mersenne.org/workload/

and log in with your account to see your assignments. Check mark the one to unreserve and then use "Unreserve checked exponents" button at the bottom.

[paulunderwood]

Originally Posted by ssybesma

Right now I am having a problem starting over.

I cleared the two jobs out of my queue (unreserved them) for this CPU (STCK1A8LFC-A079AF). I'm having trouble getting a new LL assignment.

I just want to do first checks on ONE new potential prime on this thing. The command line menu is confusing sometimes.

==========

UPDATE:

This thing is so unruly...I want to select an exponent (332375291) and it keeps sticking it under manual testing when I need it to show up under the machine that's doing the testing.

I have to have control of this and I don't yet.

I don't now much about the workings of Prime95/mprime w.r.t. PrimeNet, but https://www.mersenne.ca/exponent/332375291 says "has 100,054,933 decimal digits takes 4,944.082 GHz-days to do one L-L test". I would suggest you go with default values which PrimeNet would be happy to deal with, plus you would avoid poaching of your chosen exponent. I do understand the desire to win the prize money for the first 100M digit prime

[ssybesma]

After throwing a temper tantrum for about an hour I finally got this working properly so it looks right.

Ugh!!! It's after 4am...been up all night.

[retina]

Originally Posted by ssybesma

This thing is a complete mud-wrestling match.

It's trying to force me to do double-checks.

I am not buying 10 of these to do double-checks!!!!!!!!!!!!

Don't you want to ensure they are reliable first? No point running a large LL/PRP test and then discover they give bad results.

[ssybesma]

Originally Posted by ATH

In local.txt check for the lines WorkerThreads and CoresPerTest and change them to:
WorkerThreads=1
CoresPerTest=4

In worktodo.txt change to only 2 lines:

[Worker #1]

and then the worker line for the test you want to keep.



To unreserve the exponent on the server go to:
https://www.mersenne.org/workload/

and log in with your account to see your assignments. Check mark the one to unreserve and then use "Unreserve checked exponents" button at the bottom.

I figured out how to deal with the constant pre-assignment of numbers.

I started completely over by taking my new compute stick off GIMPS and deleting all the files it created. Then I created a worktodo.add file and put the string of information into that FROM the Manual Testing/Assignments window on the website...then when I connected to the server it put it into my worktodo.txt and fixed everything.

So now I have a completely clean looking assignment table with no manual testing going on.

The three machines I have are all working on 100 Million primes...will have 9 more doing that in the next few weeks.

Once each machine does one of those, then I'll go back and do other stuff and alternate to be a good PrimeNet citizen. I realize double checks have to be done.

[ssybesma]

Originally Posted by retina

Don't you want to ensure they are reliable first? No point running a large LL/PRP test and then discover they give bad results.

I hear what you're saying...I'll let this run for a few days.

My plan is to run this constantly for months until it completes. That is their only, dedicated purpose. My other machines won't be on those anymore once I get them all set up.

[GP2]

I can see you are running mprime version 29.4 b8

It might be better to upgrade to the latest version, 29.6 b3. You can download it from the link in this post.

Just take the mprime file from 29.6 b3 and drop it into place on top of the existing one, and then restart the program.

[Prime95]

Before you get to deep into testing, do Advanced/Time on your 332xxxxxx exponent for 100 iterations. That will give you an estimate of how long this will primality test will take. I fear an Atom processor will take more than a decade.

Should you decide to continue *please* do a PRP test, not an LL test.

[ssybesma]

Originally Posted by GP2

I can see you are running mprime version 29.4 b8

It might be better to upgrade to the latest version, 29.6 b3. You can download it from the link in this post.

Just take the mprime file from 29.6 b3 and drop it into place on top of the existing one, and then restart the program.

Thanks for the heads up on the later version!

I downloaded it and will replace the mprime executable file shortly.